Depressive illness is a common, serious condition associated with great personal suffering, significant disability, and high health care costs. While most depression treatment trials are conducted in specialty clinics, most depression treatment occurs in primary care. Previous research suggests that few patients treated in primary care receive the intensity of treatment recommended by speCialists. More recent data suggest, however, that many patients receiving low-intensity treatment have good short-term outcomes. Given that resources available to improve the treatment of depression will probably remain limited, intervention should focus on groups of patients in which more intensive treatment is likely to reduce morbidity, disability, or cost. The proposed research will focus on two dimensions of depression likely to influence clinical course, disability, and treatment response: stage of treatment (initial, recurrent, or chronic) and medical comorbidity. Longitudinal surveys of selected patient groups will assess the course of psychiatric symptoms, functional impairment, and lost productivity among patients treated for depression. Computerized data systems will be used to establish a population-based registry of all patients treated for depression in a large staff-model HMO. These computerized data will be used to assess the process of care, patterns of health care utilization, and risk of suicide for all treated patients and for specific subgroups. Data analyses will examine: * the influence of incidence, episode duration, and recurrence on overall prevalence * the prevalence and impact of comorbid medical illness * the burden of depressive illness as measured by functional impairment, lost productivity, health care costs, and suicide mortality * the process of care for patients treated with antidepressants * the relationship of process to clinical and functional outcomes Results of these analyses should help direct treatment resources to groups of depressed patients for whom more intensive treatment may reduce the overall burden of illness.